05 CHANNEL
V. SUMMARY AND CONCLUSIONS
The Caltech Electron/Isotope Spectrometer has been used to obtain the first definitive measurements of the isotopic composition of the quiet time nitrogen and oxygen cosmic rays in the region of enhancement. We have also extended the energy range of previous measurements of the elemental composition of the low energy cosmic rays.
In order to accurately identify the nitrogen and oxygen iso- topes, we calibrated a spare telescope with a variety of isotopes of 1 S Z S 9. Using Janni•s (1966) range- energy tables for protons in silicon as a basis, we calculated corrections to be applied (after scaling the tables by M2 ) for helium, nitrogen, and oxygen.
z
An enhancement, similar to that reported by other groups
(Hovestadt et al., 1973; McDonald et al., 1974; Chan and Price, 1974), was observed in the low energy nitrogen and oxygen spectra. The
dominant isotopes are 14N and 160. To the 84% confidence level the
upper limits on the 15N/N, 17
o;o,
and 18o;o
abundances are, respectively, 0.26 (5.6 - 12.7 MeV/nucleon), 0.13 (7.0- 11.8 MeV/nucleon),and 0.12 (7.0- 11.2 MeV/nucleon).
The nitrogen isotopic composition in the region of enhance- ment differs from the isotopic composition reported at higher energies. This indicates that the low energy nitrogen is not from the same population as the higher energy nitrogen. The enhancement
in the low energy nitrogen and oxygen relative to the carbon can be interpreted as a separate low energy nitrogen and oxygen component superimposed on the component decelerated from higher energies.
The dominance of 14N and 16
o
in the low energy nitrogen and oxygen cosmic rays is consistent with the theory of Fisk et al. (1974)in which neutral interstellar particles enter the solar system, are singly ionized, and are then accelerated. If, on the other hand, the particles are of galactic origin, then any proposed nucleosynthesis source must produce an isotopic composition consistent with our measurements.
APPENDIX A DATA ANALYSIS
The data processing routines relevant to this thesis will be described briefly. The purpose of this appendix is not to discuss in any depth the computer programs or the day to day data processing, but
rather to outline the general procedure. More detailed information can be found in several reports on the subject (Garrard and Hurford, 1973; Garrard, 1974b; Garrard, 1974c; Garrard and Petruncola, 1974;
Hurford, 1974; Marshall, 1974).
The flow diagram in Figure A - 1 schematically illustrates the method used to analyze the IMP-7 and IMP-8 flight data for this thesis. Briefly, the data are received in the form of experiment tapes. The experiment tapes are routinely reformatted into abstract tapes for convenience in future analysis. From the abstract tapes the events of. interest and the time averaged rates are obtained and recorded on a strip tape. For this thesis the events of interest are those that trigger 02 and 05 only, and the time interval over which rates are averaged is one day. The preceeding data analysis
is done with a PDP-ll/20 computer.
A program called HARVEST is used to analyze the strip tape on Caltech•s IBM-370/158 computer. In addition to a listing of the rate data and an event plot of the 02 channel vs. the 05 channel, punched cards are obtained listing the number of events at each 02, 05 coordinate pair. The punched output from HARVEST is then
FIGURE A - 1
Flow diagram illustrating processing sequence of flight data.
Experiment Tapes
, ,
Abstract Tapes
,,
Strip Tapes
Range - Energy Tables and Electronic Calibration
Data Events and Rates
(HARVEST)
,,
Mass and Energy ( HIZM)
,,
Mass Spectra and Energy Spectra
,,
- -
Likelihood Function(LHOOD)
,,
Isotope Abundances
submitted as input to a program called HIZM, which calculates the charge, mass, energy (corrected for the mylar window), and range of each event. The output of this program is used to obtain element abundances, energy spectra, mass spectra, etc. Knowledge of the channel-energy conversion and of the range - energy relation is required by the program HIZM. The channel - energy conversion is obtained from electronic calibration performed on the ground (see section II - C). The range - energy relation is based on calibration data for hydrogen, helium, nitrogen, and oxygen and is based on
calculation for other elements. (See Appendix
c.)
A program called LHOOO uses the mass data to calculate the likelihood function for the relative isotopic abundances.A separate, auxilliary program called t1ASCON, which is not involved in the data analysis, is used to calculate the isotope contours on a ~E-E' plot. With the channel - energy conversion and the range - energy relation as input, MASCON calculates the 02 channel and the total energy (corrected for the mylar window) as a function of the 05 channel for any isotope.
Because of the different hardware and the larger number of interesting events, a slightly different procedure is used to
analyze the calibration data. Figure A- 2 illustrates the relevant flow diagram. The pulse height data from each calibration run are recorded on one file of the PACE tape.
*
* Acronym for Pulsed Analog to digital Converter and Encoder.
FIGURE A - 2
Flow diagram illustrating processing sequence of calibration data.
PACE Tapes
,,
Events (PACBER)
~r
Mass and Energy (HIZM)
~r
Mass Histograms (MOUT)
- - - ...
02 and 05
Energy Loss
,,
Mass(D25MEV) Response
- -
Average Mass- -
,,
(MAS FIT)Differentia 1 Range Correction
(DTHICK)
,,
Range Correction
(EXCESS)
The program PACBER reads the PACE tape, calculates the number of events at each 02, 05 coordinate pair, and writes the results on tape. This tape is used as the input for the program HIZM, which calculates the charge, mass, energy,and range of each event, and creates a new tape as output. Note that since the calibrated range - energy relation is not known at this step of the analysis, other
range- energy tables are used. (See Appendix C.) Using the new tape as input, the program MOUT calculates the means and standard deviations of the mass, energy, and range distributions and also produces mass histograms, which are converted into punched cards.
Since several isotopes were obtained simultaneously for much of the calibration data, it is necessary to properly sort out the several mass peaks. Using the method of least squares, mass peaks are fitted to the mass histograms by the program MASFIT, and the result is a mean mass, a standard deviation, and a height for each mass peak. The outputs of MOUT and MASFIT can be used directly to obtain the isotope response of the instrument.
Several more steps are needed to obtain the calibrated range - energy tables. The outputs of MOUT and MASFIT are used by the program D25MEV to calculate the average energy loss in 02 and in 05 for a given charge, mean mass, and total energy. This program used the same range - energy tables that were originally used by HIZM to calculate the mass. Note that with this procedure
the isotopes have already been separated, and it is relatively easy to go back and calculate the average 02 and 05 energy loss for each isotope. It would have been more difficult to calculate the average 02 and 05 energy loss directly from the raw data, since, with an rms mass resolution of - 0.3 amu (for oxygen), it is not possible to associate the correct isotope with each event on an individual basis.
The program OTHICK uses the 02 and 05 energy loss information to calculate a differential correction {equation C - 5) to the range - energy relation. Finally, the program EXCESS calculates a range - energy correction from the differential correction.
APPENDIX B
CROSS-TALK BETWEEN DETECTORS
There is an instrument anomaly on the IMP-7 EIS whereby, under certain circumstances,a particle that deposits energy in a certain combination of detectors can also trigger the discriminator of another detector in which it did not deposit any energy (Garrard, 1974a). This phenomeon is referred to as cross-talk. There are two manifestations of cross-talk which affect 025 events on IMP-7.
Particles that deposit energy in both 02 and 05 have a certain pro- bability of also triggering 06. In this instance the 02 pulse height is replaced in the telemetry by the 06 pulse height. Such events are identified by the 0256 signature and by the zero pulse height in 06. In addition, particles that deposit a large amount of energy in 05 also trigger 04. Since 04 acts as an anticoincidence device, the event is not analyzed. 025 events on the IMP-8 EIS are not affected by cross-talk.
The cross-talk from the detector combination 025 into 06 was first observed during particle calibrations of the IMP-7 instru- ment at the Caltech Van de Graaff accelerator. A more detailed discussion than that presented here can be found in the report by Hurford (l974b).
Based on an analysis of solar flare events, the probability that a 025 hydrogen or helium event will result in cross-talk into
06 is best expressed by the equation
P(A,B) • o.s
~1
+ erf [o.1293- 81.36<\
111 -~l] l
B - 1
where P(A,B) is the cross-talk probability, erf represents the error function, and A and B represent the 05 and 02 pulse heights in units of channel number. The probability of cross-talk is greatest when both the 02 and 05 pulse heights are large, or when the 02 pulse height is - 0.9 the 05 pulse height. In order to establish a limit on the probability P(A,B) for elements with Z ~ 3, the number of 0256 events with 05 channel ~ 1500 and 06 channel
=
0 werecompared with the number of 025 events with 05 channel 2 1500. As a result, the maximum cross-talk probability was determined to be 0.40 ± .02. To calculate the cross-talk probability as a function of A and B for particles with Z~3, equation B - l is used for values of P(A,B) ~ 0.4. For all other values of A anrl B, the probability P(A,B) is set equal to 0.4.
For the nitrogen and oxygen isotopic abundance measurements of Table III - 0, the cross-talk probability given by this
technique is identical for all isotopes. For the element abundances of Table III - 3, the average cross-talk probability varies both as a function of energy and of element. The correction factors for cross-talk are listed in Table B - 1 appropriate to the elements and
Element
He
0 Li 0 Be 0
B 0
c
0
N 0
Ne 0 Mg
0
Si 0
TABLE B-1
Correction Factors for 025~06 Cross-Talk on IMP-7 EIS Energy
(MeV/nucleon) 6.0-12.7 6.0- 12.7 4.9- 12.5 4.9- 12.5 4.9- 12.5 4.9- 12.5 4.9- 12.5 4.9 - 12.5 4.9- 12.5 4.9- 12.5 5.6- 12.5 5.6 - 12.5 7.2-11.1 7.2-11.1 7.7 - 10.5 7.7- 10.5 8.1-10.2 8.1 - 10.2
Minimum Correction
Factor*
1. 67 1.14 1. 00 1.40 1. 00 1. 61 1. 00 1. 67 1. 00 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67
Maximum Correction
Factor*
1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1.67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67
Average Correction
1. 00 1.67 1. 30 1. 60 1. 56 1.60 1. 66 1. 60 1. 67 1. 60 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67 1. 67
To obtain the number of 025 events that would have been observed if there were no cross-talk, the observed number of events should be multiplied by the correction factor.
*
Calculated at the energy for which the cross-talk correction is a minimum.+Calculated at the energy for which the cross-talk correction is a maximum.
energy intervals of Table III - 3. To obtain the average cross-talk correction factor, the lithium, beryllium, and boron differential fluxes were assumed to be independent of energy. The shape of the oxygen spectrum was taken from the IMP-8 data (Figure III- 4).
The shapes of the carbon and nitrogen spectra are immaterial since the cross-talk correction factors are energy independent at these ener- gies. Since most of the cross-talk events (i.e., 0256 events with 06 pulse height = O) within the 05 channel region (channel 506 - 1187) spanned by the 6.0- 12.7 MeV/nucleon helium are due to helium particles, these events were included in the calculation of the helium flux.
Therefore, no further correction for helium cross-talk was necessary.
From Tables B - l and III - 3 the size of the cross-talk corrections can be compared with the statistical uncertainties in the element ratios. A correction for cross-talk was made in Table III - 3 only to the He/0 ratio. The N/0, Ne/0, Mg/0, and Si/0 ratios are unbiased by the cross-talk since the correction factors are the same for all elements. For the B/0 and C/0 ratios, a correction for cross-talk would be much smaller than the statistical uncertainties in the ratios. If a correction factor had been
applied to the Li/0 and Be/0 ratios of Table III - 3, then the upper limits to the ratios would have been slightly reduced. How-
ever,since only upper limits were obtained, the values in Table III - 3 are valid without a cross-talk correction.
The cross-talk from 05 into 04 was first observed during particle calibrations at the Berkeley Bevatron. This type of * cross- talk arises from the signal side of 05 being adjacent to the signal side of 04. In Figure II - the signal side of 05 is on top, and the signal side of 04 is on bottom. Because of the large detector areas and the small spacing between 04 and 05, the two detectors are capacitively coupled. Thus, a large signal in 05 can cause the 04 discriminator to trigger. For a more detailed discussion of the 05 ~ 04 cross-talk, the reader is referred to the report by Vidor (1975c).
To evaluate the 05 ~ 04 cross-talk, the IMP-7 and IMP-8 oxygen intensities were compared. The energy interval for which the comparison was made extended to the highest energy for which element abundances were calculated for the IMP-7 instrument. (See Table III-3.) Since 025 events on the IMP-8 instrument are not affected by cross-talk, the IMP-8 instrument provides a reliable measurement of the oxygen
intensity. Correcting the IMP-7 intensity for 025 + 06 cross-talk, the resulting intensities of the 5.21 - 12.74 MeV/nucleon oxygen are 4.4 ± .6 x 107 particles/sec and 3.9 ± .3 x 107 particles/sec for the
IMP-7 and IMP-8 instruments respectively. Since, at these low energies, there is no evidence of a decreased IMP-7 flux resulting from
05 ~ 04 cross-talk, no correction was made for this effect.
*
A telescope identical to that on the IMP-7 EIS was installed on the IMP-8 EIS for these calibrations.APPENDIX C
CALIBRATED ISOTOPE RESPONSE
A. Introduction
In order to calculate the mass of nitrogen and oxygen isotopes using the 6E- E' technique (see Section II-C), accurate knowledge of the range - energy relation is needed. To gain insight into the
required accuracy, we shall derive an equation that expresses the
relation between mass, range, and the measured energy loss (6E and E').
To first orde~ the range of a particle can be calculated from the known proton range (Evans, 1955). We shall express the range of the particle in a given material by
M (-ME\
R ( E, M, Z)
=
7.2 Rp i) + 6 RM,z
(E) C-1R is the range of a particle of energy E, mass M, and charge Z, where M and Z are in units of the proton mass and charge. Rp is the range of a proton of energy
N·
E 6RM,Z is a range correction for a particle of mass M and charge Z. The value of 6RM,Z is defined by equation C-1. If the stopping power~~ in a given material is a function of charge and velocity only, then the range of any isotope of element Z can be expressed in terms of the proton range and of the range correction of an isotope with mass Mz:C-2
An additional independent equation is needed to relate the mass to the measured energy loss ~E and E':
R (E,M,Z) - R (E' ,M,Z)
=
T C-3E is the total energy (~E + E') deposited in the two detectors 02 and 05, E' is the energy deposited in 05, and T is the thickness of 02.
Substituting equation C-2 into C-3 and rearranging terms, the calculated mass can be expressed as
M
=
- R p
C-4
The energies E and E' are measured quantities, and we shall assume,
for the present, that the proton range and the range correction are
known as a function of energy. For any charge Z the mass M can be
calculated in a straightforward manner using an iterative procedure.
It is interesting to note that the mass does not depend on either
range alone but only on the difference between two ranges.
Using techniques similar to the one used to calculate the mass resolution (see Section II-C), it can be shown that the relative
uncertainty in mass is- 1.3 times as great as the relative uncertainty in the range - energy relation. Therefore, to calculate the mass of an 16 0 nucleus to 0.5 amu, the range - energy relation must be known to 2.4%. By contrast, to calculate the mass of a 4He nucleus to 0.5 amu, the range - energy relation must be known to 9.6%.
In light of this need for accurate range - energy information, we review the current state of knowledge in the energy region
appropriate to this thesis. Several range - energy tables for pro- tons in silicon have been published. (E.g. Janni, 1966; Bichsel and Tschalaer, 1967). Using the large amount of experimental data as a guide, the technique for producing the range - energy tables of protons has been considerably refined. For example, Janni (1966) estimates that his tables are typically accurate to better than 1%.
In order to use the proton range - energy tables for elements with Z > 1, the range corrections ~RM Z must be known. An
z'
alternative to using proton range - energy tables and an appropriate range correction is to use directly range tables for specific isotopes.
Northcliffe and Schilling (N
&
S) (1970) have published range - energy tables for elements with 1 s: z~ 103 in a variety of media for the energy region 0.0125 ~ E s: 12 MeV/amu. Corrections were made for the effects of incomplete projectile ionization at low energies.The effects of incomplete ionization are to decrease the effective charge of the ion and thereby to reduce the stopping power and
increase the range of the particle. It should be noted that the
N
&
S range - energy tables were derived with the assumption thatthe stopping power scales with
z
2, although for the value of Z they used an effective charge based on the ionization state as a function of energy. The dependence of the stopping power onz
2arises from Bethe's formula (Livingston and Bethe, 1937), and it has been the basis of many works on the subject of ionization energy loss. N
&
S claim that the overall accuracy of {heir tables is no better than 2%, implying an uncertainty of no less than0.4 amu for the calculated 160 mass.
Recent evidence, however, has pointed to deviations from the
z
2 scaling of the stopping power which was assumed by N&
S.Several experiments have found a difference in the range of elemen- tary particles of opposite charge. Barkas et al. (1963) found a 3% difference between the range of the E + and E hyperons at identical velocities. Heckman and Lindstrom (1969) reported a similar effect for positive and negative pions. In careful experi- ments with an absolute accuracy of 0.3%, Anderson et al. (1969) compared the stopping power of hydrogen and helium in tantalum and aluminum at several MeV/nucleon. They found that the stopping power of helium relative to hydrogen was larger than the factor of 4 predicted by the Bethe formula. The discrepancy was 2.6% in
tantalum and 1.3% in aluminum at 2.5 MeV/nucleon,with the discrepancy decreasing at larger energies.
Ashley et al. (1972) and Jackson and McCarthy (1973) explained the deviations from the Bethe formula theoretically. Bethe's formula is based on the first Born approximation. By including the next higher order approximation, the authors found a term proportional to
z
3, the relative importance of which decreases with increasing velocity.Sellers et al. (1973) and Kelley et al. (1973) performed experiments which support the inclusion of the Z 3 term for 3He, 4 He, 12c, 14N, and 160 from 2 - 10 MeV/amu. The discrepancies from the Bethe formula were typically several per cent. This compares with the 2.4% accuracy needed to calculate the 16
o
mass to 0.5 amu.B. Calibration
In order to determine the isotope response as a function of energy, a spare telescope, identical to that on the IMP-8 EIS, was calibrated at the Berkeley 88" Cyclotron. The primary 12c,
14 16 .
N, and 0 beams had energ1es up to 17 MeV/nucleon. Fragmentation nuclei provided calibration data for a variety of isotopes with 1 :5:
z
:5: 9.In order to control the intensity, energy, and type of
particle seen by the telescope, the primary beam was scattered from a target. A schematic illustration of the setup is shown in Figure C-1. Several targets were used throughout the calibration: a 1.4 mg/cm2
thick gold target, tantalum targets of 11 mg/cm2,